A risk model based on signature genes predicts prognosis and associates with tumor immunity, drug sensitivity in breast cancer

BackgroundBreast cancer, the leading cause of cancer deaths among women, exhibits high heterogeneity, affecting prognosis. Understanding this heterogeneity and developing prognostic models are crucial for accurate identification of high-risk patients.MethodsAccessing breast cancer gene expression and clinical data from public datasets, we identified differential expression genes in tumor vs. non-tumor tissues using TCGA data. Key DEGs were then selected using LASSO and Cox regression, and a prognostic risk model (BRCA-DEGs-LASSO-Cox) was constructed. Survival analysis estimated model predictability, identifying high-risk patients. Correlation between risk score and signaling pathways, immune status, and drug sensitivity was analyzed. Molecular mechanisms underlying high-risk patients were discussed.ResultsOur analysis identified 1217 downregulated and 689 upregulated DEGs in breast cancer tumor tissues. A BRCA-DEGs-LASSO-Cox model was constructed using four key DEGs, stratifying patients into high/low-risk groups. High-risk patients had worse OS across cohorts and were associated with androgen, estrogen, and PI3 K signaling pathway dysregulation. They also exhibited immune status dysregulation and drug sensitivity disturbances. Molecular mechanism analysis indicated abnormal regulation of cell cycle, mitosis, and immune-related signals in high-risk patients, explaining their poorer prognosis.ConclusionsBRCA-DEGs-LASSO-Cox model effectively identifies high-risk breast cancer patients, revealing key signaling pathways, immune status, drug sensitivity, and molecular mechanisms.